Odyssey, a world model AI lab building interactive, generative video environments, has reached a $1.45 billion valuation as investors bet big on the technology underpinning the next wave of synthetic media.
Digital forensics pioneer Hany Farid warns that generative AI is making synthetic media nearly impossible to distinguish from reality, with profound consequences for trust, evidence, and detection technology.
The Atlantic built a searchable database revealing which copyrighted songs were used to train AI music generators like Suno and Udio, spotlighting the murky provenance behind synthetic audio.
Off-the-shelf AI image detectors often crumble on real-world data. Here's why generalization fails — and how DINOv2 and ConvNeXt let you train a robust, domain-specific synthetic media detector.
Snap is spinning off its AI video generation team into a new standalone company called Dotmo, citing the steep costs of building and running generative video models. The move highlights how compute-intensive AI video has become a financial burden even for large platforms.
Adobe is rolling out conversational AI assistants in Photoshop and Premiere Pro betas, letting users edit images and video through natural-language prompts — a major shift toward agentic creative workflows in synthetic media tools.
The plummeting cost of capable AI models is putting deepfake fraud tools in the hands of more criminals, raising serious risks for banks and fintech firms that rely on identity verification and biometric checks.
Midjourney, best known for generating photorealistic art, is moving into medical imaging — applying its generative expertise to synthesize full-body ultrasound scans, a striking expansion of synthetic media into healthcare.
Global AI-enabled fraud reached an estimated $442 billion last year, with voice cloning attacks now sophisticated enough to deceive trained experts and security systems alike.
The EU has published guidance on labelling AI-generated content ahead of the AI Act's August deadline, setting transparency rules that will shape how deepfakes and synthetic media are disclosed across the bloc.
New research tackles the challenge of identifying undisclosed LLM-generated content in official parliamentary texts, raising the bar for synthetic-text detection in high-stakes civic and governmental settings.
Digital forensics pioneer Hany Farid raises concerns about the reliability of automated deepfake detection tools, warning that overconfidence in these systems could undermine trust and create false security in the fight against synthetic media.